CN103517669B - The method and system of multiparameter administrative alert grade is determined during patient monitoring - Google Patents

Abstract

The present specification discloses systems and methods of patient monitoring in which multiple sensors are used to detect physiological parameters and the data from those sensors are correlated to determine whether an alarm should be issued, thereby making the alarm more precise and causing fewer false alarms. ECG readings can be combined with invasive blood pressure, non-invasive blood pressure, and/or pulse oximetry measurements to provide a more accurate picture of pulse activity and the patient's breathing. Alternatively, the monitoring system can use accelerometers or heart valve auscultation to further improve accuracy.

Description

Method and system for determining alarm levels for multi-parameter management during patient monitoring

technical field

The present specification invention relates to patient monitoring systems. In particular, the present specification discloses systems and methods for analyzing multiple physiological parameters to escalate, downgrade or suppress alarm conditions.

Background technique

Most patient monitoring is usually accomplished by measuring and observing multiple physiological parameters such as: ECG (electrocardiogram), pulse oximetry (involves measuring blood oxygen levels or SpO ₂ ), respiration (either from the ECG signal or from other parameters), invasive blood pressure (or IBP involving direct measurement of blood pressure from an indwelling catheter), and noninvasive blood pressure (or NIBP involving the use of automated oscillometric methods).

Typically, these physiological parameters have a set of vital signs and derived measurements that can be configured to alert caregivers if the measurements move outside configured ranges. Each parameter has various alerts that can be considered with different priorities. However, prior art methods and systems tend to treat each of these parameters independently when determining/determining an alarm condition, or are unable to effectively determine that an alarm condition derived from a particular patient monitoring system is false, and potentially false , or sufficient instructions to ensure that the patient's status provides a viable mechanism for alerting the nursing staff. As a result, clinical users may experience an unacceptable number of alarms within these patient monitoring systems. Nurses end up seeing a dense alarm state from the various ups and downs of each parameter, leading to unnecessary distraction and indifference to the alarm by the caregiver.

Thus, there is a need in the art for a method and system that effectively suppresses or downgrades the number of false alarms seen by the user and ensures that when the system alarms, there is a high probability that immediate patient attention is required.

Contents of the invention

In one embodiment, the specification discloses a computer-readable medium storing a plurality of program instructions for processing data indicative of a physiological parameter, comprising: a) code for receiving ECG data generated at least in part by an ECG device, wherein the ECG data comprising a plurality of features and wherein at least one of said features has a designation associated therewith and a time of occurrence associated therewith; b) receiving a code indicative of the pulse data of the patient pulse response, wherein said pulse data is obtained from at least one sensor separate from said ECG device, and wherein said pulse data has a designation associated therewith and a time of occurrence associated therewith; c) combining said at least one sensor of ECG data a designation and time of a characteristic correlated with a designation and time of the pulse data to determine a degree of correlation; and d) code for causing an alarm to be issued only if the degree of correlation indicates that the patient has an abnormal cardiac condition.

Optionally, the plurality of program instructions further includes code for comparing the degree of association with a predetermined value. The code that causes the alarm to sound the alarm only if the comparison indicates that the patient has an abnormal heart condition. The designation of at least one characteristic of the ECG data is normal or abnormal. The designation of this pulse data is normal or abnormal. This correlation serves to determine whether abnormal features in the ECG data are temporally associated with abnormal pulses. If the correlation determines that abnormal features in the ECG data are temporally associated with abnormal pulses, an alert indicating an abnormal cardiac condition is issued. If not, do not raise the alert, or actively suppress the alert if generated by another source. The correlation is further dependent on at least one of an amplitude of the ECG signal, an amplitude of the pulse signal, a duration of the pulse signal, a noise level within the ECG data, or a noise level within the pulse data. The at least one sensor is an invasive blood pressure (IBP) monitoring device, a non-invasive blood pressure (NIBP) monitoring device, a heart valve sound monitoring device, or a pulse oximetry (SpO ₂ ) monitoring device. The plurality of instructions further includes code for causing at least one sensor to initiate collection of pulse data based on the ECG data. The code causes the non-invasive blood pressure monitoring device to inflate the cuff and collect pulse data when the ECG data represents a heart rhythm indicative of atrial fibrillation. The plurality of instructions further includes code for causing the noninvasive blood pressure monitoring device to inflate a cuff and collect pulse data based on the association.

In another embodiment, the specification discloses a computer-readable medium storing a plurality of program instructions for processing data indicative of a physiological parameter, the plurality of program instructions comprising: a) receiving biological data generated at least in part by a respiratory monitoring device; code for impedance data, wherein said generated impedance data comprises a plurality of features and wherein at least one of said features has a designation associated therewith and an occurrence time associated therewith; b) receiving respiration data indicative of patient respiration code, wherein the respiration data is obtained from at least one sensor separate from the respiration monitoring device, and wherein the respiration data has a designation associated therewith and a time of occurrence associated therewith; c) converting the respiration impedance data to Designation and timing of said at least one characteristic of said at least one feature is correlated with designation and timing of respiratory data to determine a degree of correlation; and d) code for causing an alarm to be issued, wherein said degree of correlation indicates that the patient has abnormal breathing alarm.

Optionally, the respiratory monitoring device is at least one of a carbon dioxide concentration monitoring device, a pneumatic respiratory transduction device, a strain gauge or an extensometer. The sensor is at least one of an ECG device, an invasive blood pressure (IBP) monitoring device, a pulse oximetry (SpO ₂ ) monitoring device, or a motion detection device. The motion detection device is an accelerometer. The motion detection device is an accelerometer integrated with ECG electrodes. The designation of at least one characteristic of the bioimpedance data is normal or abnormal. The designation of the respiration data is normal or abnormal. This correlation serves to determine whether abnormal features in the bioimpedance data are temporally correlated with abnormal respiration data. If the correlation determines that anomalous features in the bioimpedance data are temporally associated with abnormal respiration data, an alert indicative of a respiration condition is issued. The breathing condition is a sleep apnea event. The plurality of instructions further includes code for receiving motion data from the accelerometer and determining whether the patient has fallen. The plurality of instructions further includes code for receiving motion data from the accelerometer, determining whether the patient is engaging in an activity that increases the patient's respiration rate, and causing the alert to sound based at least in part on the determination. The plurality of instructions further comprises receiving motion data from the accelerometer, receiving ECG data, determining whether a change in the ST segment of the ECG data was caused by patient activity, and causing or not sounding the alarm based at least in part on the determination code.

It should be understood that the instructions described herein are stored in a storage structure such as a hard disk, ROM, RAM, or any other type of storage device for execution by at least one processor. The command can be co-located with the sensors or monitors or remote from them. They can be integrated into a separate controller or computer in data communication with the sensors, or work as software modules integrated into one or more sensing devices themselves.

Description of drawings

These and other features and advantages of the present invention will become better understood with reference to the following detailed description when considered in conjunction with the accompanying drawings, in which:

Figure 1 shows a flowchart depicting a method of determining an alert level using a plurality of parameters;

Figure 2a is a graphical representation of an ECG (III) signal in the presence of waveform noise and the corresponding IBP (ART) signal showing normal cardiac activity over the same interval;

Figure 2b is a graphical representation of an ECG (III) signal in the presence of waveform noise and the corresponding IBP (ART) signal showing a lack of cardiac activity in the same interval;

Figure 3a is a graphical representation showing plethysmography-assisted analysis of an atrial ectopic beat;

Figure 3b is a graphical representation of ECG (V1), ECG2 (II) and _SpO2 signals at time _T0 with plethysmography-assisted ventricular beat analysis;

Figure 3c is _a graphical representation of ECG (V1), ECG2 (II) and _SpO2 signals at time T1 with plethysmography-assisted ECG signal noise analysis;

Figure 4 is a graphical representation of invasive blood pressure systolic peak modulation using a respiratory signal;

Figure 5a is an illustration of one embodiment of a disposable ECG electrode;

Figure 5b is an illustration of the disposable ECG electrode shown in Figure 5a, further showing the reusable snap-on electrode wire;

Figure 5c is an illustration of the reusable snap-on wire shown in Figure 5b, attached to an ECG electrode and with an integrated accelerometer;

Figure 6 is a graphical representation of chest wall motion in a supine patient wearing an accelerometer;

Figure 7 is a graphical representation of the chest wall motion after standing up of the same patient wearing an accelerometer; and

Figure 8 is a graphical representation of the chest wall motion of the same patient wearing an accelerometer while walking.

detailed description

In one embodiment, the present specification discloses systems and methods for centrally analyzing multiple physiological parameters and using the results to escalate, downgrade or suppress alert notifications. The present specification provides the benefits of generating more specific patient alarms and reducing the occurrence of false alarms so that monitoring personnel can perform more efficiently.

In one embodiment, ECG parameters are considered together with one or a combination of the following sensor measurements: invasive blood pressure (IBP); non-invasive blood pressure (NIBP); and blood oxygen levels such as via pulse oximetry (SpO ₂ ). For each parameter there is a corresponding waveform signal generated by measuring and sampling the signal leaving the transducer.

For ECG, the waveform is derived from the electrical signal detected by subcutaneously placed electrodes in response to the propagation of the electrical signal in the heart muscle. In one embodiment, the IBP uses an indwelling catheter with a transducer to generate a voltage proportional to the pressure from the mechanical pumping action of the heart.

NIBP measurements are obtained via an external cuff coupled to an electronic pressure transducer. The cuff automatically inflates and deflates at intervals to measure pressure oscillations. While NIBP is used to measure blood pressure, the pulse rate is also typically determined and reported as part of that process. For example, a caregiver may set up or program the monitor to take NIBP measurements every 15 minutes. This is typical in the operating room (OR) or in the post-anesthesia care unit (PACU) setting. When taken every 15 minutes, NIBP measurements will report something like "120/80(92) HR77" (i.e., systolic = 120 mmHg, diastolic = 80 mmHg, mean arterial = 92 mmHg, and Pulse rate = 77bpm). In this case, the NIBP parameters essentially provide an independent measure of pulse rate, but only do so every 15 minutes.

In another embodiment, for purposes of this specification, the cuff is inflated periodically, such as every few minutes, to a pressure sufficient to measure pulse rate. In one embodiment, the cuff is inflated to a diastolic pressure equal to or slightly greater than the most recently measured diastolic pressure. In another embodiment, the cuff is inflated to a mean arterial pressure equal to or slightly greater than the most recently measured mean arterial pressure. In another embodiment, the cuff is inflated so that both the diastolic and mean arterial pressures are equal to or slightly greater than the most recently measured corresponding pressures. Pulses detected upon cuff inflation were used as a surrogate source of pulse information in the same manner as described for IBP and _SpO2 .

In yet another embodiment, NIBP is used to measure the strength and regularity of the pulse signal in addition to the pulse rate.

The _SpO2 waveform is derived by measuring changes in the amount of light detected by photoreceptors after light has passed through the patient's skin. The anatomical site used must have arterial blood flow through it in sufficient quantity, like a fingertip or an ear.

In any case, for each parameter a signal corresponding to the electrical activity on the heart or the pumping action of the heart and its subsequent propagation to the periphery of the body is generated. Each parameter is provided to the caregiver to verify the results obtained via the electrical signal (ECG) collected on the skin and as a pulse signal via the invasive pressure line (IBP), external cuff (NIBP), or pulse oximeter (SpO ₂ ) Independent implications of agreement between measured mechanical responses.

Further, when the patient is initially monitored, each waveform is processed independently to derive a record of the occurrence of each event (beat or pulse), and a number of parameters for each event are measured and recorded. For each ECG event (i.e., heartbeat), the system measures and records the waveform deflection across multiple leads and multiple recordings if the deflection pattern is typical, and if it falls on the expected place in the sequence based on previous events. height and direction. Additionally, other factors like duration, rate of change, and locations of local minima and maxima within each lead are recorded. Ultimately, all recorded measurements are combined and compared to previous beats, and a diagnosis is made as to whether the beat represents a "normal" or "abnormal" condition.

For purposes of this specification, determining whether an alarm should be raised is the time of the occurrence of the ECG signal (typically marking a salient feature of the ECG waveform), the designation of the ECG signal as "normal" or "abnormal", and the estimate of the system's confidence in beat diagnostics function. If the beat is normal in all measured parameters (closely matching previous beats), occurs at the expected time, and all other measures of signal consistency and quality are high, then the system will have this signal as Reliable High A confidence level, eg, a confidence level that exceeds a predetermined threshold. All characteristics measured by the ECG signal processing algorithm are reported to the signal correlation software module, which then processes the data to issue a confidence level and compares the confidence level to a threshold. Similarly, other waveform parameters including time of occurrence, amplitude, duration, rate of change of peak, and measurements of signal quality (such as, but not limited to, IBP, NIBP, or _SpO2 ) are recorded and reported to the signal correlation software module.

Combine the measured characteristic data from each parameter using the signal correlation software module. Under normal conditions, each electrical pulse as measured by ECG produces an impulse response that is also measured among other parameters. The relationship between occurrence time, pulse duration, noise level, and confidence is established over time. When the signal quality is good, each ECG complex captures a good mechanical response in the heart, and every other parameter produces a good impulse response, the agreement or correlation between each parameter is very high.

In one embodiment, when an abnormal beat (early or late, atrial or ventricular ectopic) is detected via ECG, there is a high likelihood of a decrease in impulse response in one of the other parameters. If such ectopic beats occur at a certain frequency, a pattern develops between the ECG detecting an "abnormal" condition and a reduced impulse response, among other parameters. This pattern is recognized by the system as presenting a high level of confidence that it represents a real event, triggering an alert.

In one embodiment, other parameters report a normal impulse response while the ECG signal is affected by noise (usually a result of patient motion), and reports an "abnormal" condition. In this embodiment, the condition is actually "normal", but the ECG signal is masked by noise. Information from other parameters (a good and consistent pulse signal with high confidence in the expected time) is used to suppress any alarms or notifications about abnormal beats. The ECG then uses the information gleaned from other parameters to fully reconsider its diagnosis. Similarly, feedback from the pulse source can help downgrade or suppress high and low pulse rate alarms and asystolic alarms caused by signal quality issues on the ECG electrodes. This is a consequence of the high correlation previously established between ECG and pulse sources. When the data from the pulse source is of high quality and produces expected results, the system can suppress or downgrade the alarm from the ECG source.

Conversely, when an actual event like an asystolic pause (heart stops beating) occurs, the ECG will detect and report no activity and the pulse source will detect and report no impulse response. All of these parameters together produce a closely related signal and consider the heart to have stopped. The system then triggers the alert with the highest urgency.

The present invention is intended to provide multiple embodiments. The following disclosure is provided to enable those of ordinary skill in the art to practice the invention. Language used in the specification should not be construed as a disavowal of generality to any one particular embodiment, or used to limit the claims beyond the meaning of the terms used therein. The general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the invention. Furthermore, the terms and phrases used are for the purpose of describing the exemplary embodiments and should not be regarded as limiting. Accordingly, the present invention should be accorded the widest scope encompassing numerous alternatives, modifications, and equivalents consistent with the principles and features disclosed. For the purpose of clarity, details relating to technical matters that are known in the technical fields to which the invention pertains have not been described in detail so as not to unnecessarily obscure the present invention.

FIG. 1 shows a flowchart depicting a method of analyzing a number of parameters according to the importance of the alarm to establish an alarm level, thereby determining whether to provide an alarm (via an audio or visual signal) to a caregiver. In one embodiment of the multi-parameter alarm grading method of the present invention, IBP (invasive blood pressure), NIBP (non-invasive blood pressure), and _SpO2 (blood oxygen levels like via pulse oximetry) sensor measurements are combined to Consider ECG parameters. First, measure/record (105) parameters like time of occurrence, signal strength, amplitude, and regularity of each pulse signal recorded by the IBP, NIBP, and/or _SpO2 sensors. Thereafter, in step 110, a one-to-one correlation is established between each measured ECG complex and the resulting pulse signal as measured for IBP, NIBP, and/or _SpO2 .

In step 115, correlations between pulse sources (ie, IBP, NIBP, and/or _SpO2 sensors) are continuously monitored and analyzed. After due consideration is given to the composite pulse rate reading from the IBP, NIBP, and/or _SpO2 sensors and the ECG, a heart rate alarm condition is determined in step 120 in order to increase the overall confidence level of the alarm condition. If the overall confidence level is high, as when parameters from multiple sources are coordinated, then in step 125 an alarm is sounded or escalated. However, if the confidence level is low, as is the case when parameters from multiple relevant sources are inconsistent with each other, the alert is suppressed or downgraded in step 130 .

In one embodiment, for example, using the above-described method of the present invention, by simultaneously observing pulse signals from an invasive blood pressure sensor, a cuff blood pressure sensor, and/or a pulse oximeter, and when sufficient confidence in the pulse signal is such that Detect and suppress false ECG arrhythmia alerts when suppressing ECG-based alerts. Therefore, if there is a sufficiently strong rhythmic pulse signal as measured on the IBP, NIBP, _SpO2 sensors, there is a reasonable certainty that the patient is not experiencing an arrhythmia. In such instances, ECG alarms will be de-escalated or suppressed at the level at which the alarm is audible to the caregiver in accordance with the method of the present invention, thereby avoiding problems associated with symptoms such as asystole, ventricular tachycardia, ventricular beats, and ventricular tachycardia. False alarms associated with arrhythmic conditions such as paroxysmal tachycardia. Similarly, an ECG arrhythmia alarm is escalated if the arrhythmia condition is confirmed or corroborated by information from the source of the IBP, NIBP, and/or _SpO2 pulse. For example, ectopic beats are often selected for lower pulse pressure and smaller blood flow. This reduced peripheral pressure or flow can be detected on the _Sp02 signal, external cuff, and/or arterial pressure line. The presence of diminished signal in _Sp02 , the cuff, and/or the crimp confirms or increases the confidence in labeling those beats as ectopic.

Figure 2a is a graphical representation of an ECG(III) signal 200 in the presence of waveform noise 205 and a corresponding IBP(ART) signal 201 showing normal cardiac activity 210 over the same interval. This graph shows that the noise 205 closely resembles the ECG waveform 200 of ventricular tachycardia, which typically generates a high priority alarm. However, because the simultaneous IBP waveform 201 clearly shows a continuous pulse 210 with a very regular rhythm and amplitude, this high priority alarm was downgraded to a low priority alarm indicating a "noisy ECG". Figure 2b is a graphical representation of an ECG(III) signal 202 in the presence of waveform noise 220 and a corresponding IBP(ART) signal 203 showing a lack of cardiac activity 225 within the same interval. This graph shows a segment of ventricular tachycardia 220 identified by cessation of pulsatile activity 225 in the invasive blood pressure waveform. A high priority alert was escalated due to the large degree of correlation between the signals obtained from the two independent measurements.

Figure 3a is a graphical representation showing plethysmography-assisted analysis of an atrial ectopic beat. As can be observed in the waveform of Figure 3a, a weak _Sp02 plethysmographic signal 305 confirms the ectopic beat 310 in the ECG waveform.

Figure 3b is a graphical representation of the ECG(V1) 311, ECG2(II) 312, and SpO ₂ 313 signals at time _T0 with plethysmography-assisted ventricular beat analysis, and Figure 3c is the plethysmographic _- assisted ECG signal noise analysis at time T1 Graphical representation of ECG (Vl) 321 , ECG2 (II) 322 and SpO ₂ 323 signals. Referring now to FIG. 3 b , a ventricular beat V is confirmed or confirmed in the SpO ₂ graph 313 correspondingly showing a small or no signal response 315 . This confirmation increases the overall confidence level, thereby escalating the heart rate alarm status. However, since the ECG data becomes noisy at the end of band 320 shown in Figure 3b, and the beginning of band 325 shown in Figure 3c, the associated _SpO2 waveforms give the possibility that neither of this noise is true Pulsating confirmation. This makes it possible to downgrade or suppress otherwise audible heart rate alarms if the _Sp02 waveform and ECG heart rate parameters are relied upon independently.

In another embodiment, a NIBP (non-invasive blood pressure) cuff is used as an alternative signal source in the same manner as IBP and/or _SpO2 when necessary. The system inflates the cuff periodically and when a basal heart rate source (ECG, IBP, SpO ₂ ) is unavailable, inconsistent, or indicates a serious condition that needs to be verified. The system will perform periodic inflation as well as on-demand inflation as described below.

For example, in one embodiment, the patient is monitored for NIBP and ECG only. A patient accidentally removes some or all of their ECG electrodes, rendering the ECG parameters useless. At this point, the cuff is inflated and pulse monitoring begins with NIBP "backup". If the NIBP produces a reasonable "in-bounds" pulse signal, the system communicates a low-priority alert ("check ECG leads" or "signal not available") to the nursing staff.

However, if the pulse rate indicates an alarm condition (like no pulse, high rate, low rate, or very different pulse strength or regularity than previously measured), then raise the alarm message to something like "ECG not available, NIBP indicates pulse rate >120bpm—Clinical alarm like "check patient".

In another example, if the ECG analysis deems the rhythm change to be atrial fibrillation, the NIBP cuff is inflated on the patient to monitor the strength and regularity of the pulses measured in the NIBP cuff. This additional NIBP data is then used to confirm or suppress an atrial fibrillation diagnosis. Similarly, out-of-bounds heart rate alarms are checked via the NIBP cuff to confirm or deny rate violations before sounding the alarm.

It should be noted by those of ordinary skill in the art that various other combinations of parameters may be utilized in accordance with the alarm level determination method of the present invention, and that the use of parameters like ECG, IBP, NIBP, and/or _SpO2 is by way of non-limiting example. example to illustrate. However, it should be noted that precise suppression or escalation of alarms requires careful correlation of different physiological parameters to ensure tracking and reporting of real events. It should also be noted here that the occurrence of ECG-derived events is still reported to patient monitors by notification or display, rather than as any visible or audible signal that could be designed to draw attention and indicate abnormal physiology requiring immediate medical attention. status of clinical alerts. This enables caregivers to review excluded events from the multiparameter analysis of the present invention.

According to another aspect of the present invention, multiple cross-physiological parameters are also analyzed to determine the alarm level. One of ordinary skill in the art will appreciate that depending on the placement of the catheter, the hemodynamic condition of the patient, and the characteristics of the respiration, a respiration signal can be derived from the invasive blood pressure line. In one embodiment, an invasive blood pressure line signal, observable as a breath-driven pulse pressure change, is used as the auxiliary breathing signal. Therefore, the invasive blood pressure line signal is used in conjunction with the basic respiration signal to confirm changes in respiration rate or to identify apneic events.

Figure 4 is a graphical representation of invasive blood pressure systolic peak modulation using a respiratory signal. Thus, FIG. 4 shows how the signal curve of the systolic peak 405 of invasive blood pressure is modulated with a respiration signal 410 in accordance with an embodiment of the present invention. This modulation is measured and used as an auxiliary source for respiration rate measurements. As can be observed in Figure 4, the relationship between the bioimpedance respiration signal 410 and the invasive blood pressure signal 405 is established during the first six seconds. As the bioimpedance signal 410 deteriorates, at 415 the blood pressure signal 405 is used to establish the respiration rate despite the momentary unavailability of the bioimpedance signal 410 , thereby suppressing false respiration rate related alarms. However, the fact that a bioimpedance signal was not available was still reported as an event rather than as a clinical alert.

In one embodiment, the assist breathing signal is derived by monitoring changes in the amplitude of the ECG signal in a plurality of leads during the breathing cycle. These amplitude changes are a result of the movement of the heart in the chest relative to the measuring electrodes due to the movement of the chest and lungs during respiration. This, in turn, creates a problem when combined with a basic source of respiratory signals like a capnometer, a PRT (pneumatic respiratory transducer), a strain or extensometer, or any source of bioimpedance signals known to those of ordinary skill in the art. Another pseudo respiration signal to confirm respiration changes when studying/analyzing. Since this artifact is not expected, and therefore always unreliable (since in this example the artifact depends on chest/lung motion), only when a high correlation is observed between the artifact and the underlying respiratory signal This pseudo-signal is only used when the

In another embodiment, the assisted breathing signal is derived by monitoring changes in the amplitude of the _SpO2 plethysmogram or small changes in the oxygen saturation signal during the breathing cycle. These amplitude changes are a result of the movement of the heart in the chest due to the movement of the chest and lungs during respiration, thus creating another false respiration signal that is used to confirm respiration changes when studied/analyzed in conjunction with the underlying source of the respiration signal .

In one embodiment, a motion signal from a motion detecting accelerometer is used in combination with a primary source of a respiration signal as a pseudo or auxiliary respiration signal. In one embodiment of the invention, the accelerometer is integrated into the wire snap, making the accelerometer almost reusable at very low one-time cost.

Figure 5a is an illustration of one embodiment of a disposable ECG electrode, while Figure 5b is an illustration of the disposable ECG electrode shown in Figure 5a, further showing a reusable snap wire.

In one embodiment of the invention, an accelerometer (not shown) is integrated into the wire snap 505 . In one embodiment, the electrode wire is attached to the ECG electrode 500 as illustrated in Figure 5c. In one embodiment, a three-axis accelerometer is used, such as, but not limited to, the ADXL3303-axis from Analog Devices Accelerometer. Those of ordinary skill in the art will appreciate that accelerometers can be integrated into other devices than ECG electrodes like cellular phones, or other devices like iPod ^™ from Apple ^™ .

In one embodiment, the spatial orientation of the accelerometer is maintained so that it is consistently applied in the same orientation relative to the patient. To do this, mechanical fasteners like locking snaps, joints, or adhesives are used to consistently position, align, and lock the accelerometer snaps into one position. In one embodiment, a marker like a sticky note that says "this end up" is placed on the electrode/accelerometer along with a locking tab or any other mechanical fixation to ensure that the accelerometer is always oriented the same relative to the body , and stay in this position. The accelerometer-integrated ECG electrodes are properly positioned on the patient so that the chest wall motion is maximized. Those of ordinary skill in the art will appreciate that the accelerometer can be used integrally with the ECG electrodes and/or independently. In one embodiment, two or more different 3-axis accelerometers are used, placed at different locations on the patient's torso for maximum detection of the measured physical quantity.

In conditions where only the accelerometer signal is available and other basic respiration signals are not available or available, the accelerometer data is used as a proxy "dead in bed" detector that sounds an alarm when no motion or respiration signal is present. However, in situations where another source of the basic respiratory signal is available (eg, bioimpedance, stress or strain gauges, capnography), the accelerometer signal is used to verify the basic respiratory signal. Thus, when the accelerometer measurements agree with those of the other underlying sources, there is increased confidence that the measured signal is correct and an audible alert is given to the caregiver. However, data from the proxy accelerometer signal is used to suppress false alarms from other sources of respiration when the signal from the accelerometer is indicative of a different respiration signal with a high enough confidence such a situation). Therefore, data from multiple sources are jointly analyzed to derive more robust respiration measurements. This improves the quality of respiratory signal analysis and reduces unnecessary alarms for the caregiver.

According to one aspect of the invention, motion signals from accelerometers are used to determine and monitor patient posture. For example, counting and reporting the amount of time a patient spends in standing, sitting, active and/or supine positions.

Figure 6 is a graphical representation of chest wall motion in a supine patient wearing an accelerometer. A 3-axis accelerometer measures applied force along each of 3 different orthogonal directions. Figure 6 depicts a graphical representation of the signal generated by the chest-worn accelerometer while the patient is lying down. The data represents a 1 minute period showing the signals from each axis of the accelerometer. The effect of breathing can be seen in FIG. 6 , particularly in waveform G 605 and waveform Y 610 , and to a lesser extent, waveform B 615 . The average signal for waveform G605 is approximately 300 counts, the average signal for waveform Y610 is approximately -150 counts, and the average signal for waveform B615 is approximately -350 counts.

Figure 7 is a graphical representation of the chest wall motion after standing up for the same patient wearing an accelerometer. Thus, changes in patient position lead to significant changes in the average level of each signal. Compared to the patient supine, the average level of waveform G705 is now -130 counts, the average level of waveform Y710 is -650 counts, and the average level of waveform B715 is 150 counts. This effect is the result of a change in the spatial orientation of the 3-axis accelerometer relative to the Earth's gravitational field. Provided that the accelerometer is attached to the patient in the same orientation each time, the average of the signals off each axis will tell the clinician whether the patient is "upright" (which can be standing or sitting), supine, or if in an upright May be partially supported by pillows from the supine position.

From a monitoring point of view this information can be used in evaluating excretion decisions. Similarly, sleeping positions like back, left side, right side, abdomen can be monitored. In one embodiment, the number of position changes per hour is measured and quantified and used in conjunction with motion measurements from the accelerometer to determine whether the patient's overall activity is what the patient is expected to do in a given state. For example, the detection of lack of movement or changes in posture can be notified to caregivers, if needed, so that the patient can be turned over in bed to prevent bedsores.

In another embodiment, the data collected by the accelerometer representing the change in position is used to detect a fall and trigger an alarm notification. For example, a waveform feature representing a sudden change from a standing position to a supine position prior to impact would alert the system that the patient may have fallen. The system then escalates or triggers an alarm notification.

According to another aspect of the invention, in addition to monitoring patient posture, accelerometer signals are used to measure patient activity. In one embodiment, accelerometer signals are used to count and record patient steps and patient step rate.

Figure 8 is a graphical representation of the chest wall motion of the same patient wearing an accelerometer while walking. The time axis represents 10 seconds of data collected while the patient was walking. The walking characteristics are quite different from the supine and standing characteristics depicted in Figures 6 and 7, respectively. The user can easily make out each footstep (especially in waveform Y810 and waveform G805) and see approximately 19 steps taken in 10 seconds. Actual pace is measured by looking for characteristic rapid changes in each axis signal across all 3 bands. These events can be counted (pedometer type function) and reported as counts or as a rate; for example, in this ten second period, the events can be expressed as a total of 19 steps or 114 steps per minute.

This information is used to calculate statistics like the percentage of time the patient spends walking and at what stride rate. Such statistics, when analyzed over a long period of hours and days, in one embodiment, help in assessing parameters such as the following for ambulatory patients: How much can they walk? How do their activity levels compare to similar patients? Are they candidates for excretion? And/or any other parameter as advantageously would be apparent to a person of ordinary skill in the art.

According to another aspect of the invention, accelerometer signals are used to detect signal changes caused by motion/activity artifacts among other physiological parameters being monitored on a patient. In one embodiment, this may be used in suppressing or de-escalating alarm conditions during periods of heightened patient activity. For example, respiration measurements by methods like bioimpedance and pulse oximetry can be severely compromised by rough patient movement. Those of ordinary skill in the art will appreciate that ECG signals can be impaired by motion. To assess motion-induced artifacts, analysis of the motion signal from the accelerometer is used to correlate specific noise in the ECG signal with specific patient motion/activity including walking. Such a sufficiently positive correlation enables false ECG alarms to be suppressed. In another embodiment, the use of accelerometer signals in combination with other physiological parameters can increase the confidence level of an alarm condition. For example, a possible marginal arrhythmia detected on the ECG accompanied by a motion signal consistent with fainting is escalated to a higher alarm priority.

In another embodiment, changes in patient posture detected using accelerometers are used to analyze and interpret measured changes in ECGST segments. As is well known to those of ordinary skill in the art, the ST segment is the portion of the ECG waveform that is monitored to identify an ongoing myocardial infarction. Sometimes ST segment levels vary with patient position as a result of the movement of the heart within the chest relative to the ECG electrodes. However, changes in the ST segment caused by positional changes were not significant. The accelerometer signal provides the necessary information to convey the ST-segment change prior to the position change, thereby enabling the de-escalation of the ST-segment change alarm.

According to one aspect of the invention, the motion signal from the accelerometer is used to modify the overall sensitivity of the patient alarm system. In one embodiment, if the accelerometer signal analysis strongly suggests that the patient is walking or very active, then the sensitivity of the alarm system in the monitor is desensitized appropriately. The goal here is to reduce false alarms caused by patient activity. According to one aspect of the invention, patient activity levels monitored using accelerometer signals are used to alter the need and type of analysis performed on other physiological parameters. For example, if the patient's activity level is high due to activities such as walking or using a treadmill, ECG analysis of the patient is interrupted during the activity because ECG analysis requires high signal quality. Instead, in one embodiment, the overall sensitivity level of the alarm system is reduced so that motion-induced noise signals do not trigger false alarms while some fundamental parameter analysis continues.

According to yet another aspect of the invention, heart valve sounds are measured (eg, microphone placed on the chest) to monitor mechanical activity of the heart, thereby improving overall patient monitoring and reducing false alarms.

In one embodiment, heart valve sounds are used as a measure of the patient's pulse activity. The valve sounds from the heart form separate pulse signals that are used to distinguish noise from the signal on the ECG electrodes. As a first step, valve sound signatures matching each QRS location detected on the ECG were identified and recorded. In a next step, the quality of the recorded valve sound signal is determined on a beat-by-beat basis according to parameters like intensity, consistency, quality and regularity of the sound signal. The determined sound quality measure or signal-to-noise ratio is continuously evaluated. The resulting valve sound quality is then used to weigh how strongly data from the sound channel (like a microphone placed on the patient's chest) is used to escalate or suppress alarm data from other physiological parameter measurement channels/sources like ECG electrodes .

In another embodiment, valve sounds from the heart form separate pulse signals used to identify non-perfusing beats and pulseless electrical activity. Those of ordinary skill in the art will appreciate that pulseless electrical activity is a general condition of electromechanical dissociation in the heart. In some arrhythmic situations, it is advantageous to identify beats without a mechanical response. In other words, these beats are nonperfusing because they have an electrical signal (perhaps abnormal) but do not cause the heart to pump. For example, there are cases where cardiac pacemakers induce electrical signals detected by ECG parameters in the heart, but do not produce effective mechanical pumping. In this case, the mechanical response by way of the valve sound signal measured beat-by-beat combined with the ECG signal enables such events and alarms to be properly identified. For example, if an arrhythmia event detected on the ECG is verified by valve sound signals, the probability that this is a real event and that this alarm can be escalated increases due to increased confidence that the alarm is correct. Such a situation exists, for example, if the ECG analysis considers pause or asystole (no beat detected) and the heart valve audible signal also considers no mechanical movement. This is the case where the diagnosis of pause or no contraction is confirmed, with confidence escalating this alert. Similarly, if an event is detected by the ECG but not indicated by the heart valve audible signal, the event alarm is suppressed or downgraded. For example, an ECG signal may indicate a train of irregular beats thought to be ventricular tachycardia. However, the high-quality heart valve sound signals that the pulse rate does not match the irregular beat that the ECG analysis is detecting. Therefore, in this case, the alarm of the irregular beat is suppressed or downgraded.

Claims (25)

1. A system for processing data indicative of a physiological parameter comprising: means for receiving ECG data generated at least in part by an ECG device, wherein the ECG data includes a plurality of features, at least one of the features has a designation associated therewith and a time of occurrence associated therewith, and the ECG data indicative of a ventricular tachycardia condition; means for receiving impulse data indicative of a patient's impulse response, wherein the impulse data is obtained from at least one sensor separate from the ECG device, and wherein the impulse data has a designation associated therewith and a Appearance time; means for associating the designation and timing of the at least one characteristic of the ECG data with the designation and timing of the pulse data to determine a degree of correlation; and means for causing a low priority alert indicating a noisy ECG or a high priority alert, wherein the low priority alert is issued when said pulse data associated with said ECG data indicates a regular rhythm and amplitude, wherein a high priority alert is issued when said pulse data associated with said ECG data indicates cessation of pulsation. 2. The system of claim 1, further comprising means for comparing the degree of association with a predetermined value. 3. The system of claim 2, wherein the means for causing a high priority alert causes the alert only if the comparison indicates that the patient has an abnormal cardiac condition. 4. The system of claim 1, wherein the designation of at least one characteristic of the ECG data is normal or abnormal. 5. The system of claim 4, wherein said designation of the pulse data is normal or abnormal. 6. The system of claim 1, wherein the correlation functions to determine whether an abnormal feature in the ECG data is temporally correlated with an abnormal pulse. 7. The system of claim 1, wherein if the correlation determines that an abnormal feature in the ECG data is temporally associated with an abnormal pulse, an alert indicating an abnormal cardiac condition is issued. 8. The system of claim 1 , wherein the correlation is further dependent on the amplitude of the ECG signal, the amplitude of the pulse signal, the duration of the pulse signal, the level of noise within the ECG data, or the noise level within the pulse data. At least one of the noise levels. 9. The system of claim 1, wherein the at least one sensor is an invasive blood pressure (IBP) monitoring device, a non-invasive blood pressure (NIBP) monitoring device, a heart valve sound monitoring device, or a pulse oximetry (SpO ₂ ) monitoring device equipment. 10. The system of claim 1, further comprising means for causing at least one sensor to initiate collection of pulse data based on the ECG data. 11. The system of claim 10, wherein said means for causing at least one sensor to initiate collection of pulse data from said ECG data causes a non-invasive blood pressure monitoring device to monitor a cuff when said ECG data represents a cardiac rhythm indicative of atrial fibrillation. The bag is inflated and pulse data is collected. 12. The system of claim 1, further comprising means for causing a non-invasive blood pressure monitoring device to inflate a cuff and collect pulse data based on said association. 13. A system for processing data indicative of a physiological parameter, comprising: means for receiving bioimpedance data generated at least in part by a respiratory monitoring device, wherein said generated impedance data comprises a plurality of features, and wherein at least one of said features has a designation associated therewith and a time of occurrence associated therewith ; means for receiving respiration data indicative of a patient's respiration, wherein the respiration data is obtained from at least one sensor separate from the respiration monitoring device, and wherein the respiration data has a designation associated therewith and an occurrence associated therewith time; means for associating the designation and timing of the at least one characteristic of the bioimpedance data with the designation and timing of the respiration data to determine a degree of correlation, wherein if bioimpedance data is not available, invasive blood pressure systolic peak data is used instead of bioimpedance data; and means for enabling event reporting or alerting, wherein the event reporting is only issued when bioimpedance data is not available but the correlation using invasive blood pressure peak systolic data indicates that the patient is breathing normally, Wherein the alarm is issued only when the degree of correlation indicates that the patient has abnormal breathing. 14. The system of claim 13, wherein the respiratory monitoring device is at least one of a capnography monitoring device, a pneumatic respiratory transduction device, a strain gauge, or an extensometer. 15. The system of claim 14, wherein the sensor is at least one of an ECG device, an invasive blood pressure (IBP) monitoring device, a pulse oximetry ( _SpO2 ) monitoring device, or a motion detection device. 16. The system of claim 15, wherein the motion detection device is an accelerometer. 17. The system of claim 15, wherein the motion detection device is an accelerometer integrated with ECG electrodes. 18. The system of claim 13, wherein the designation of at least one characteristic of the bioimpedance data is normal or abnormal. 19. The system of claim 18, wherein the designation of the respiration data is normal or abnormal. 20. The system of claim 19, wherein the correlation functions to determine whether abnormal features in the bioimpedance data are temporally correlated with abnormal respiration data. 21. The system of claim 20, wherein if the correlating determines that anomalous features in the bioimpedance data are temporally associated with abnormal respiration data, an alert indicating a respiration condition is issued. 22. The system of claim 21, wherein the respiratory condition is a sleep apnea event. 23. The system of claim 16, further comprising means for receiving motion data from the accelerometer and determining whether a patient has fallen. 24. The system of claim 16, further comprising means for receiving motion data from the accelerometer, determining whether the patient is engaging in an activity that increases the patient's respiration rate, and causing the alert to sound based at least in part on the determination. 25. The system of claim 16, further comprising receiving motion data from the accelerometer, receiving ECG data, determining whether a change in the ST segment of the ECG data is caused by patient activity, and based at least in part on the determination using The component that the alarm is issued or not issued.